This application describes a 5-year training program of didactic study, research mentoring and research activities focusing on the genetics of childhood obsessive-compulsive disorder (OCD). The training program will allow the candidate to gain expertise in methods of genetic research in complex disorders, including linkage, association and bioinformatics. The long-term objective of the career development plan is to extend the candidate's expertise in psychiatric phenotyping, statistical genetics and bioinformatics. To accomplish this goal, research activities will use the resources of the Johns Hopkins Department of Psychiatry for the understanding of the OCD phenotype; the Hopkins School of Public Health and National Human Genome Research Institute for bioinformatics, statistical genetics and genetic epidemiology training; and the Hopkins Genetics Core Lab facility for genotyping. Preliminary work includes the candidate's previous experience in a NIMH-funded family study of OCD with adult probands and pilot feasibility data from OCD child proband families. The research design consists of 2-stage sampling and psychiatric interviews of 80 OCD child probands and 720 first-, second-degree relatives. Probands are recruited by consecutive nonprobability sampling to maximize their utility for linkage and segregation. Only non-sib relatives undergo screening for greater efficiency of phenotyping. The research plan components are: a) phenotypic characterization of childhood OCD; b) linkage using affected-relative pairs to 4 candidate gene regions (SLC6A4-serotonin transporter, COMT, MAOA, DRD4) using 3 SNP biallelic markers and 4 flanking microsatellite markers; c) complex segregation analysis to obtain the mode of transmission and major gene effect; d) family-based tests of linkage disequilibrium using the 4 candidate gene alleles and; e) tests of genetic heterogeneity with the tic-related OCD subtype. Based on results from initial analyses and developing collaborations with other research groups future research projects, such as a genome-wide screen of OCD, are considered.